The marRAB multiple antibiotic resistance operon of Escherichia coli controls the expression of a large number of genes resulting in low level antibiotic and superoxide resistance through a complex network of reactions. MarR auto-represses the mar operon but is inactivated upon interaction with salicylate, losing its DNA binding capacity. This, in turn, results in derepression of the operon and expression of MarA, which activates the transcription of some 40 to 60 promoters including the marRAB promoter itself (auto-activation). The activation of the target (regulon) genes by salicylate is determined by several factors including the binding affinity of the particular promoters to MarA and the activation of a subset of the genes of the regulon (e.g. fumC and sodA) by a second, as yet uncharacterized, mar-independent pathway. Detailed analyses of the salicylate induction are being carried out in collaboration with Dr. Michael Wall (Los Alamos National Laboratory) in an attempt to develop a mathematical model for the induction network. Two mechanisms by which MarA interacts with RNA polymerase (RNAP) to activate promoters of the regulon are under investigation. We previously identified (in collaboration with Drs. Bindi Dangi and Angela Groneneborn) the interacting surfaces of MarA and of the carboxy-terminal domain of the alpha subunit of RNAP (alpha-CTD) by NMR-based chemical shift mapping and demonstrated that this interaction is fundamentally distinct from that used by alpha-CTD to interact with other transcriptional activators such as CRP and Fis. The alpha-CTD interaction is necessary for the activation of all promoters of the regulon. A second interaction is also required for about half of the promoters of the regulon ? those promoters (Class II) in which the MarA binding site overlaps the ?35 binding site of RNAP. , This interaction was expected to involve the sigma-70 subunit of RNAP and to be common to all Class II promoters, based on studies with other AraC proteins (of which MarA is a member). We have now carried out extensive analyses with mutants of MarA to determine which surface amino acids might interact with sigma-70 and find that certain amino acids have a promoter-specific interaction with sigma 70. In addition, further evidence for which amino acids are involved in MarA-sigma-70 interactions has come from a collaboration with Dr. Ann Hochshild (Harvard Medical School). Using her sigma-70?alpha-NTD chimeras we have corroborated the importance of certain amino acids in the interaction of s sigma-70 with MarA. We are now attempting to obtain direct structural evidence concerning the interaction of MarA with sigma-70 at different promoters. In an effort to identify further members of the mar regulon, we have recently found that salicylate induces the stringent response (collaboration with Drs. David Grainger and Steve Busby, The University of Birmingham). The same CHIP analysis found that one operon of E. coli, yhcRQ, is strongly induced by salicylate in a mar-independent manner as the result of activation by YhcS, a member of the lysR family of transcriptional activators. Furthermore, this operon is unique in that antibody to sigma-70 immunoprecipitated mRNA from the entire operon suggesting that sigma-70 remains with the RNAP throughout transcription. We have cloned the appropriate regions of the E. coli chromosome and are currently analyzing the transcriptional regulation of this operon. A problem of considerable theoretical interest is whether it is possible to identify promoter sequences from sequence data alone. We are using both experimental and statistical approaches to attack this problem. Because of their high degree of degeneracy RNAP binding sites should occur with far greater frequency than the number of genes in the E. coli chromosome. To investigate such potential promoters, we randomly cloned small fragments (~165 bp) of DNA into an expression vector and analyzed them for promoter function. We found weak promoter activity at a frequency corresponding to one promoter every 107 bp. Among the most highly active of such promoters detected on a plasmid, in situ chromosomal transcripts corresponding to more than half of the sequences tested could be identified by RT-PCR. However, no stable mRNA was found for any of these fragments although each appeared to contain sequences homologous to the consensus RNAP binding signals. This suggests that the mRNA produced from these promoters is unstable. Factors identified for converting these fragments into efficient promoters included: a ribosome binding site and an UP element. ("Detection of low level promoter activity within ORF sequences of Escherichia coli" Mitsuoki Kawano, Gisela Storz, B. Sridhar Rao, Judah L. Rosner and Robert G. Martin, Nucleic Acids Research, accepted for publication.) A second statistical approach has begun with identification of sequences in E. coli that match the RNAP binding signals (TTGACA and TATAAT) in at least 10 of the 12 positions and are spaced 16, 17 or 18 nt apart. The 625 sequences found are not randomly distributed since 200 lie immediately upstream of ORFs or stable RNA operons and another 127 lie in intercistronic regions. That is, over half are not within ORFs although ORFs cover 80 to 90% of the genome. We are currently analyzing the sequences immediately up- and downstream of these near-perfect RNAP consensus sequences in an effort to determine whether some sequence element (other than their position within an ORF) can distinguish those that serve as promoters and those that do not. In addition, in collaboration with Drs, Julio Collado-Vides and Araceli Huerta-Moreno, we are analyzing whether those near-perfect sequences that do serve as promoters have evolved differently from those that do not. This work was carried out in collaboration principally with Dr. J.L. Rosner and with Angela M Gronenborn, PhD (SB, LCP, DIR, NIDDK), Mitsuoki Kawano, PhD (EGR, CBMB, DIR, NICHD), and Gisela Storz, PhD (EGR, CBMB, DIR, NICHD)